For this proposed study, follow-up assessments will be conducted at ages 12 to 17 of a group of predominantly white chidren whose mothers were seen when the children were 5 to 10 years of age. The original sample consisted of 583 mothers. We anticipate that 85 percent will participate at T2. Previously, extensive data were collected from mothers with regard to the children's personality characteristics, behavioral problems, their own personality attributes, their childrearing practices, and their drug use. Their offspring will also participate in the T2 data collection providing descriptions of their drug use. Structured inverviews will be conducted separately with the mother and the adolescent and the interview schedules will include a number of scales with adequate psychometric properties. Data obtained in the proposed current assessments will be related to data obtained earlier so that the following may be investigated: (a) the association of early risk fact ors with later drug use; (b) the continuity-discontinuity characterizein different areas of the youngster's development and its association with the adolescent's drug use; and (c) the in-depth background of later correlates of adolescent drug use. Our conceptualizations are rooted in the disciplines of epidemiology, psychology, and psychiatry. Our primary analytic technique will be either hierarchical multiple regression of structural equation causal analysis. The major contribution of the proposed study will be to provide data on the early determinants of adolescent drug use, an area about which very little is known at the present time. This information is necessary not only for the prediction of drug abuse but for its prevention and treatment as well.